The obesity epidemic in the U.S. continues unabated, and is accompanied by substantial complications, including diabetes, cardiovascular disease, and death. Obesity is strongly influenced by sequence variation in as yet unknown susceptibility genes. Whole genome association scans offer a potentially powerful method for identifying common variants with modest effects on obesity, but the deluge of data from these studies will require careful analysis and rigorous follow-up. The first whole genome association studies are now being performed in large samples in which body mass index has been measured, and we will have access to whole genome association data from 3,000 individuals with measures of obesity (body mass index, BMI). We hypothesize that these studies will identify many potential associations between genetic variants and obesity, representing a mix of largely false positive results interspersed with a smaller number of true associations. It will be critical to perform appropriate and rigorous follow-up to distinguish the true causal variants from the false leads. The combination of multiple large well-characterized cohorts with measures of obesity, high throughput genotyping and robust analytic methods will permit us to rapidly follow up the preliminary results from the whole genome association scans. Specifically, we will be able to extract from the mass of data those genetic variants that are truly associated with obesity, even if the effects are modest. This will lay the groundwork for future studies of the phenotypic consequences of these variants on other obesity-related phenotypes and the risks of complications such as diabetes and cardiovascular disease, and for studies of gene-gene and gene-environment interactions. Successful identification of genes that are convincingly associated with obesity would highlight key pathways that influence obesity in humans, guiding efforts at therapy and prevention. [unreadable] [unreadable] [unreadable]